REVIEW Open Access

Development, functional organization, andevolution of vertebrate axial motor circuitsKristen P. D’Elia and Jeremy S. Dasen*

Abstract

Neuronal control of muscles associated with the central body axis is an ancient and essential function of the nervoussystems of most animal species. Throughout the course of vertebrate evolution, motor circuits dedicated to control ofaxial muscle have undergone significant changes in their roles within the motor system. In most fish species, axialcircuits are critical for coordinating muscle activation sequences essential for locomotion and play important roles inpostural correction. In tetrapods, axial circuits have evolved unique functions essential to terrestrial life, includingmaintaining spinal alignment and breathing. Despite the diverse roles of axial neural circuits in motor behaviors, thegenetic programs underlying their assembly are poorly understood. In this review, we describe recent studies that haveshed light on the development of axial motor circuits and compare and contrast the strategies used to wire theseneural networks in aquatic and terrestrial vertebrate species.

Keywords: Motor neuron, Spinal circuit, Neural circuit, Development, Evolution, Axial muscle

BackgroundThe neuromuscular system of axial skeleton plays crucialroles in basic motor functions essential to vertebrates,including locomotion, breathing, posture and balance.While significant progress has been made in decipheringthe wiring and function of neural circuits governing limbcontrol [1, 2], the neural circuits associated with axialmuscles have been relatively under studied, particularlyin mammals. Despite comprising more than half of allskeletal muscles in mammals, how axial neural circuitsare assembled during development is poorly understood.Although all vertebrates share similar types of axial

muscle [3, 4], the nervous systems of aquatic and terres-trial species control these muscle groups in distinct ways.In most aquatic vertebrates, rhythmic contraction of axialmuscle is essential for generating propulsive force duringswimming, the predominant form of locomotion used byfish. In land vertebrates, axial circuits have been largelydissociated from locomotor functions, and have beenmodified throughout evolution to enable new types ofmotor capabilities. In animals with upright postures, neur-onal control of axial muscles is essential to maintain bal-ance and proper alignment of the spine. During the

invasion of land by vertebrates, axial muscles that wereinitially used in swimming were also adapted by the re-spiratory system to enable breathing in air. Since many ofthese diverse axial muscle-driven motor behaviors areencoded by neural circuits assembled during develop-ment, insights into the evolution of axial circuits mightemerge through comparisons of the genetic programs thatcontrol neural circuit assembly in different animal species.In this review, we discuss studies which have investi-

gated the development, evolution, and wiring of neuronalcircuits essential for control of axial muscle. Recent ad-vances in genetically tractable systems, such as zebrafishand mouse, have provided novel insights into the mecha-nisms through which axial circuits are assembled duringdevelopment, and have shed light on the wiring of the cir-cuits essential for balance, breathing, and locomotion. Wecompare the strategies through which animals generatedistinct classes of spinal neurons that coordinate axialmuscles, with particular focus on the spinal motor neuronsubtypes that facilitate axial-driven motor behaviors.

Functional organization and peripheralconnectivity of axial motor neuronsAlthough used for fundamentally distinct motor functions,the axial neuromuscular systems of fish and tetrapodsshare many anatomic features and early developmental

* Correspondence: jeremy.dasen@nyumc.orgNeuroscience Institute, Department of Neuroscience and Physiology, NYUSchool of Medicine, New York, NY 10016, USA

© The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

D’Elia and Dasen Neural Development (2018) 13:10 https://doi.org/10.1186/s13064-018-0108-7

programs [3, 4]. In both fish and tetrapods, axial musclescan be broadly divided into two groups, epaxial andhypaxial, which are initially separated by a horizontalmyoseptum (Fig. 1a). Epaxial muscles reside dorsal to themyoseptum and include muscle groups associated withthe vertebral column and base of the skull. Hypaxial mus-cles are predominantly located ventral to the mysoseptumand give rise to diverse muscle groups including abdom-inal and intercostal muscles, as well as the diaphragm inmammals. In tetrapods, migratory populations of hypaxialmuscle also generate all of the muscle in the limb. In fishand amphibians, the separation between dorsal and ven-tral axial muscles is maintained in adulthood, while in tet-rapods many of these positional differences have been

lost. Both types of axial muscles receive innervation fromspinal motor neurons (MNs) and sensory neurons thatproject either along the dorsal (epaxial) or ventral (hypax-ial) branches of the spinal nerves.In tetrapods, MNs targeting specific muscle groups are

organized in discrete clusters, termed motor columns andmotor pools [5–8]. Spinal MNs projecting to functionallyrelated muscle groups, such as epaxial, hypaxial, or limbmuscle, are contained within motor columns that occupyspecific rostrocaudal positions within the spinal cord.Within these columnar groups, MNs further segregateinto motor pools, each pool targeting a single muscle.Each pool occupies a specific position within the spinalcord, and its relative position along the dorsoventral,

a b

c d

Fig. 1 Organization of axial MNs in tetrapods and fish. a In jawed vertebrates, axial muscles are separated into dorsal epaxial and ventral hypaxialgroups, separated by the horizontal myoseptum (HM). Each muscle group is innervated by separate spinal nerves. Dorsal root ganglia (drg) andsympathetic chain ganglia (scg) are shown. b MN columnar subtypes at trunk levels. In tetrapods, as well as some cartilaginous fish, MNsinnervating dorsal epaxial muscles are organized in the medial motor column (MMC). MNs projecting to ventral hypaxial muscles are containedwithin the hypaxial motor column (HMC). Autonomic preganglionic column (PGC) neurons, which project to scg, are shown in gray. c Organization ofMN pools at thoracic levels. MNs innervating specific types of axial muscle are organized in pool-like clusters. Some MNs within the HMC project todorsally located axial muscles, such as serratus, but are nevertheless supplied by axons originating from the ventral ramus. Abbreviations: tv,transversospinalis; long, longissimus; ilio, iliocostalis; lc, levator costae; sr, caudal serratus; ii, internal intercostal; sc, subcostalis; ei, externalintercostal; eo, external oblique. Not all trunk muscles are shown. Diagram based on data from rat in [13]. d Organization of MNs in adultzebrafish. MNs innervating fast, intermediate, and slow muscle are organized along the dorsoventral axis. Fast MNs include primary MNsand some secondary MNs, intermediate and slow are all secondary MNs. These MN types project to specific types of trunk-level axialmuscles. Diagram based on data in [14]

D’Elia and Dasen Neural Development (2018) 13:10 Page 2 of 12

mediolateral, and rostrocaudal axes is linked to how MNsproject within a target region. The stereotypic organizationof MN position within the spinal cord therefore establishesa central topographic map which relates neuronal settlingposition to target specificity.Studies on the developmental mechanisms controlling

MN columnar and pool organization have largely focusedon the diverse subtypes innervating limb muscles [9, 10].Axial MNs also display a topographic organization that re-lates neuronal position to target specificity. The cell bodiesof MNs targeting epaxial and hypaxial muscles are orga-nized in specific columnar groups within the ventral spinalcord (Fig. 1b). Dorsal epaxial muscles are innervated byMNs in the median motor column (MMC), while hypaxialmuscles are innervated by MNs in the hypaxial motor col-umn (HMC). MMC neurons occupy the most medial pos-ition of all spinal MNs, whereas HMC neurons, and allother MN subtypes, typically reside more laterally [11]. Likelimb MNs, both MMC and HMC neurons further differen-tiate into specific pool groups, and axial MN pool positionis linked to the location of its muscle target (Fig. 1c). Forexample, MMC neurons targeting more dorsal epaxialmuscles reside more medially than those targeting moreventral muscle [12]. A similar somatotopic organization hasbeen observed for HMC pools targeting different intercos-tal and abdominal muscles [13].In contrast to tetrapods, the organization of axial MNs

into well-defined columnar groups has not been de-scribed in zebrafish. Despite the absence of an obviouscolumnar organization, zebrafish axial MNs are func-tionally organized along the dorsoventral axis of thespinal cord (Fig. 1d). This organization is associatedwith how MNs are recruited at distinct swimmingspeeds and correlated with the type of muscle a MNinnervates, as opposed to the location of the muscle.Axial MNs projecting to muscles activated at slowswimming speeds reside ventrally, MNs recruited atfast swimming speeds are located dorsally, and MNsinvolved in intermediate speeds sit between fast andslow MNs [14–16].Although a clustered organization of axial MN has not

been described in zebrafish, in certain cartilaginous fishspecies, including the little skate and catshark, the cell bod-ies of MMC neurons are clustered and settle in a ventralposition [17]. These observations suggest that theorganization of axial MNs into columns was present in thecommon ancestor to cartilaginous fish and tetrapods, andtherefore to all jawed vertebrates with paired appendages.Notably, unlike most fish species, skates do not use theaxial muscles to generate propulsive force during locomo-tion, which is instead provided by contraction of the pec-toral and pelvic fins. The organization of MNs intocolumnar and pool groups therefore does not appear tohave evolved with terrestrial locomotion, but rather reflects

differences that emerged between certain fish species andother vertebrate classes.

Genetic programs specifying early axial motorneuron fatesHow are the distinct identities of MMC and HMC neu-rons established during tetrapod development? As withother subtypes of spinal MNs, the progenitors that giverise to axial MNs are specified through secreted signal-ing molecules acting along the dorsoventral axis of theneural tube shortly after its closure [18]. These morpho-gens establish specific molecular identities through theinduction of transcription factors in neuronal progenitors,which subsequently specify the identity of each of the majorclasses of spinal neuron. In the ventral spinal cord, gradedShh signaling induces expression of transcription factorswhich specify MN and ventral interneuron progenitor iden-tities [19]. As progenitors differentiate, additional transcrip-tion factors are expressed within postmitotic cells and actto define specific neuronal class fates [20]. Spinal MN pro-genitors are derived from a domain characterized by ex-pression of Olig2, Nkx6.1, and Pax6. As postmitotic MNsemerge, they initially express the Lim homeodomain pro-teins Islet1, Islet2 (Isl1/2), Lhx3, Lhx4 (Lhx3/4), as well asthe Mnx-class protein Hb9 (Fig. 2a).As MNs differentiate and migrate to their final settling

positions, subtypes of axial MNs can defined by differen-tial expression of Lim HD and Mnx factors [11, 21]. Intetrapods, MMC neurons maintain expression of Hb9,Isl1/2, and Lhx3/4, whereas the majority of other MNsubtypes, including HMC neurons, downregulate Lhx3as they become postmitotic (Fig. 2b). The specific func-tions of Lhx3 and Lhx4 in MMC neurons are not com-pletely understood, as both genes are required for thedifferentiation of all spinal MN subtypes [22]. Neverthe-less, misexpression of Lhx3 can convert limb MNs to anMMC fate and redirect motor axons towards axialmuscle, indicating that Lhx3 plays an instructive role indetermining the trajectories of MMC motor axons to-wards epaxial muscle [23]. While trunk-level HMC neu-rons can be also defined by expression of specifictranscription factor combinations, whether these factorsare required for columnar-specific differentiation pro-grams is currently unknown.A key step in the specification of axially-projecting MNs

is the segregation of newly born neurons into MMC andHMC subtypes. MMC neurons are thought to representthe ancestral “groundstate” of MNs from which all othersubtypes subsequently evolved [24]. This idea is supportedby the observation that MMC identity is the default differ-entiation state of MNs derived from embryonic stem cells(ESCs) generated through induction with retinoic-acidand Shh [25, 26]. In addition, MMC-like neurons drivelocomotor behaviors in limbless vertebrates such as the

D’Elia and Dasen Neural Development (2018) 13:10 Page 3 of 12

lamprey and insect larvae, suggesting that an MMC-likeMN population represents the ancestral condition of MNsin bilaterians.In tetrapods, an obligate step in MMC differentiation

is the sustained expression of Lhx3/4 in post-mitoticMNs; while in HMC neurons and all other MN subtypesLhx3/4 must be downregulated for proper differentiation[21, 23]. The maintenance of Lhx3/4 in MMC neuronsappears to be partially governed by Wnt signaling origin-ating from near the floorplate of the spinal cord (Fig. 2a)[27]. Overexpression of Wnt4 or Wnt5a promotes thespecification of MMC neurons at the expense of otherMN subtypes in chick embryos, while combined geneticremoval of Wnt4, Wnt5a, and Wnt5b in mice leads to de-pletion in MMC number. Recent studies in ES cell-derived MNs suggest that additional signaling pathwaysact in conjunction with Wnt signaling to promote MMCspecification [28]. Inhibition of Notch signaling in ES-cellderived MNs promotes the specification of HMC neuronsat the expense of MMC neurons, suggesting that Wnt4/5and Notch cooperate to specify MMC identity.While the extrinsic and intrinsic factors governing the

specification of MMC and HMC neurons have beencharacterized, the downstream effectors of their fate de-terminants are less well-understood. Soon after leaving

the cell cycle, the axons of MMC and HMC neuronsbegin to project outside the spinal cord, both initiallypursuing ventrolateral trajectories. The axons of MMCneurons separate from the main nerve and extend dor-sally, while all other MN subtypes, including HMC neu-rons, continue to extend ventrolaterally. The dorsaltrajectory of MMC neurons appears to rely on target-derived chemoattractant signaling emanating from asomite-derived structure, the dermomyotome [29, 30].This region expresses fibroblast growth factors (FGFs)which act on the axons of MMC neurons that selectivelyexpress FGF receptor 1 (Fgfr1) (Fig. 2b) [31]. Mutation ofFfgr1 in mice causes defects in the peripheral trajectory ofMMC axons. In addition, misexpression of Lhx3 leads toectopic expression of Fgfr1 in non-MMC MNs and causeslimb motor axons to gain sensitivity to FGFs [31].

Specification of axial MNs in zebrafishIn zebrafish, spinal MNs innervating axial muscle are spe-cified by the same core groups of transcription factors thatact in tetrapods. Unlike amniotes, where all MNs are gen-erated during a single wave of neurogenesis, zebrafishhave two waves of MN birth, primary and secondary. Pri-mary and secondary neurons are each important for dif-ferent types of axial muscle-based behaviors, but not

a

b c

Fig. 2 Specification of axial MNs in tetrapods and fish. a Specification of early axial MN identities. Graded sonic hedgehog (Shh) acts along thedorso (d)-ventral (v) axis to specify MN progenitors (pMN) and ventral interneuron fates. Graded Wnt signaling promotes sustained expression ofLhx3 in MMC neurons, while Hox signaling specifies segmentally-restricted MN columnar fates, including limb-innervating lateral motor column(LMC) neurons. b Axial MNs in tetrapods can be defined by expression of specific transcription factors. MMC neurons express Fgr1 and areattracted to mesodermally-derived FGF signaling. c Primary MNs in zebrafish. Four distinct axial MN types can be defined by their rostrocaudalposition and muscle target specificity. dRoP, dorsal rostral primary; vRoP, ventral rostral primary; CaP, caudal primary; MiP, middle primary MN

D’Elia and Dasen Neural Development (2018) 13:10 Page 4 of 12

distinguished by any known transcription factor [32, 33].Primary MNs, which number three to four per hemi-segment, are born between 10 and 14 hours post-fertilization (hpf), develop subtype-specific electricalmembrane properties as early as 17 hpf, and begin axoninitiation at 17 hpf [34, 35]. Although one or two commonMN markers such as Isl1, Isl2, and Mnx proteins can helpdifferentiate two or three primary MN subtypes at differ-ent ages, these factors cannot distinguish them through-out development and have dynamic expression patternsthat make the subtypes challenging to track over time[36–38]. All early-born MNs require the Olig2 transcrip-tion factor [39], while Nkx6 proteins appear to be requiredonly in a subset of primary MNs [40]. Postmitotic primaryMNs can be defined by differential expression of Mnx/Hb9, Isl1/2, and Lhx3 factors [37, 38, 41–43].Most genetic studies of axial MN specification in zeb-

rafish have largely focused on the specification of thefour major types of primary MNs: the dorsal rostral pri-mary (dRoP), ventral rostral primary (vRoP), caudal pri-mary (CaP), and middle primary (MiP) subtypes (Fig.2c). dRoP and MiP MNs are similar to MMC neurons,in that they project to muscles located dorsal to thehorizontal myoseptum, while CaP and vRoP project ven-trally. However, unlike MMC and HMC neurons in tet-rapods, these primary MN types cannot be distinguishedby differential expression of Lhx3. Nevertheless, disrup-tion of the core MN determinants Lhx3/4, Isl1/2, andMnx leads to defects in primary MN specification andconnectivity. For example, loss of Lhx3/4 leads to MNswith hybrid MN/interneuron fates [41], while loss ofMnx proteins affects the specification of MiP MNs [38].While much is known about primary axial MNs, the

later-born secondary MNs have been particularly under-studied. Although they make up the majority of spinalMNs in zebrafish, and are thought to be more similar tomammalian MNs, very little is known about their differ-entiation programs [44]. Secondary MNs are born start-ing at 16 hpf, begin axon initiation at 26 hpf, and areproduced to an undetermined time after 25 hpf [35].Multiple studies have described up to ten different axial-muscle innervating subtypes, six of those are secondaryMNs [45]. All MN subtypes can be differentiated basedon birthdate, muscle target, soma size and position,presence or absence of intraspinal or intermyotomal col-laterals, and firing properties. There are three distincttypes of firing patterns expressed by zebrafish axial MNsat 4 dpf: tonic, chattering, and burst firing. Tonic firingpatterns are specific to primary MNs, while chatteringand burst firing patterns are specific to secondary MNs.Each secondary MN subtype has a different distributionof these two firing patterns. While the distinct physio-logic and anatomic features of secondary MNs have beenwell-characterized, it is yet unknown whether they

reflect the operation of MN-intrinsic genetic programsacting during development.

Diversification of tetrapod axial motor columnsWhile axial MNs of fish and mammals share severalcommon early developmental programs, in tetrapodsthese subtypes have undergone a significant degree ofmodification throughout the course of vertebrate evolu-tion. All of the segmentally restricted subtypes of spinalMNs, including the diverse MN populations innervatinglimb muscle, appear to have evolved from the ventrally-projecting HMC-like population. This hypothesis is sup-ported by the observation that in genetic mutants withdisrupted specification of non-axial MN subtypes, af-fected populations revert to a predominantly HMC-likemolecular profile. Genetic deletion of the limb MN fatedeterminant Foxp1 in mice causes a loss of limb-specificMN programs and an expansion in the number of MNswith an HMC-like molecular identity [21, 46]. Expres-sion of Foxp1 in limb-innervating lateral motor column(LMC) neurons is governed by Hox transcription factorsexpressed at specific rostrocaudal levels of the spinalcord, and Hox genes are essential for generating the di-verse motor pool populations targeting specific limbmuscles [47–49]. MMC neurons appear to be insensitiveto the activities of Hox proteins, likely due to the func-tionally dominant actions of Lhx3 [21, 23]. The diversifi-cation of tetrapod spinal MNs appears to stem fromHMC-like precursors which co-opted Hox genes to gen-erate more specialized populations.Hox-dependent regulatory programs also contributed

to the diversification of MNs targeting specific hypaxialmuscle types. An important step in the evolution ofmammals was the appearance of a novel MN subtypededicated to control of respiratory muscles. MNs innerv-ating the diaphragm are contained within the phrenicmotor column (PMC) and require the actions of twoHox genes (Hoxa5 and Hoxc5) for their specification[50]. Similar to the role of Foxp1 in limb MNs, loss ofHox5 genes disrupts PMC specification programs anddiaphragm innervation, with the remaining MNs revert-ing to a thoracic HMC-like identity (Fig. 3a, b). As aconsequence, mice lacking Hox5 genes show severe de-fects in respiratory function and perish at birth [50, 51].Hox5 proteins act in conjunction with more MN-restricted fate determinants, including the POU-classhomeodomain protein Scip (Pou3f1), which is also es-sential for respiratory function [52]. Downstream targetsof Hox5 and Scip activities include genes encoding thecell adhesion proteins Cdh10 and Pcdh10, which appearto be important for PMC neurons to cluster into colum-nar groups [53].Whether MMC neurons targeting specific epaxial

muscles show the same degree of molecular diversity as

D’Elia and Dasen Neural Development (2018) 13:10 Page 5 of 12

HMC-derived MNs is less clear. While all MMC neu-rons can be defined by the maintenance of Lhx3/4 ex-pression, the specific determinants of MMC subtype-specific properties are poorly defined. A recent study in-vestigating the function of Pbx transcription factors inspinal MN differentiation identified a novel repertoire ofgenes selectively expressed in mature MMC neurons[54]. Pbx proteins are known to be important cofactorsfor Hox proteins, and are essential for the specificationof segmentally-restricted neuronal subtypes [55]. Muta-tion of Pbx genes in spinal MNs disrupts the specifica-tion of all Hox-dependent subtypes, with the majority ofthe remaining MNs consisting of MMC and HMC neu-rons. Surprisingly, removal of Pbx genes also leads a lossof the somatotopic organization of the remaining Hox-independent MMC and HMC populations. In Pbx mu-tants, MNs with MMC and HMC molecular identitiesare generated at all rostrocaudal spinal levels, but MNsof each type are randomly distributed within the ventralcord (Fig. 3c).Loss of Pbx genes does not affect the ability of MMC

and HMC neurons to select appropriate muscle targets[54], suggesting a specific function of Pbx targets in gov-erning MN columnar organization. Gene targets actingdownstream of Pbx proteins are therefore essential for theability of axial MNs to coalesce into specific columnargroups. Identification of genes differentially expressed be-tween normal and Pbx mutant MNs uncovered a novelrepertoire of targets that are selectively expressed inMMC neurons (Fig. 3d). These downstream targets

include the transcription factor Mecom (MDS1/Evi1),which marks postmitotic axial MNs and can be inducedby forced misexpression of Lhx3 in non-MMC popula-tions. The disorganization of axial MNs in Pbx mutantstherefore appears to be a consequence of the disruption ofregulatory programs acting in MMC neurons.

Development of locomotor axial motor circuits infishWhile the connections established between axial MNsand muscle play important roles in shaping motor func-tions, how the activities of different classes of MNs arecontrolled during specific motor behaviors are less well-understood. Activation of specific MN subtypes is or-chestrated through the inputs they receive from higherorder “premotor” microcircuits within the spinal cordand brain. In many cases, these premotor networks as-semble into rhythmically active central pattern generator(CPG) networks to control basic behaviors such as walk-ing, swimming, and breathing [1, 56, 57]. Much of ourunderstanding of the functional and electrophysiologicalproperties of CPG networks stem from studies of axialmuscle-driven motor circuits in the lamprey, which definedthe core neuronal constituents of CPGs [58]. Recent studiesin genetically tractable systems, such as zebrafish, havedrawn new attention to the role of axial MNs in shapingfunctional properties of locomotor CPG networks.The first movements of the embryonic zebrafish begin

at 17 hpf with altering coil contractions of the trunk thatincrease in frequency until 19 hpf and decrease until 27

a b

c d

Fig. 3 Diversification of axial MN subtypes in tetrapods. a At rostral cervical levels, HMC-like precursors give rise to phrenic motor column (PMC)neurons through the actions of Hoxa5 and Hoxc5 proteins. The activities of Hox5 proteins are inhibited by Lhx3 in MMC neurons, and Foxp1 inLMC neurons. Hox5 proteins work in conjunction with the Pou domain protein Scip to promote PMC-restricted gene expression. b In the absenceof Hox5 genes, PMC neurons are disorganized and revert to an HMC-like state. c Pbx genes are required for the columnar organization of axialMNs. In the absence of Pbx genes, Hox-dependent MN subtypes (LMC and PGC neurons) are lost, and acquire an HMC fate. The remaining HMCand MMC subtypes are disorganized at all spinal levels. d Pbx proteins act in conjunction with other MMC-restricted factors such as Lhx3 to promoteMMC specific gene expression

D’Elia and Dasen Neural Development (2018) 13:10 Page 6 of 12

hpf [32]. These early spontaneous coiling contractions inthe embryo are not dependent on synaptic transmission,but involve electrically coupled networks of a subset ofpremotor interneurons that are rhythmically active anddependent on gap junctions [33]. Ipsilateral neurons areelectrically coupled and active simultaneously, whilecontralateral neurons are alternatively active [33]. At 21hpf, zebrafish will partially coil in response to touch and,at 27 hpf, zebrafish will swim in response to touch. Thesetouch responses, and swimming thereafter, depend on glu-tamaterigic and glycinergic chemical synaptic drive anddescending inputs from the hindbrain [32, 33]. Propulsionduring swimming is generated by alternating, neural-mediated waves of muscle contractions along the trunk ofthe fish.The organization of MNs in the zebrafish spinal cord

correlates with their functional role. This relationship isbecause the MNs are grouped according to which typeof muscle fiber they innervate (Fig. 1d) [14]. For ex-ample, the dorsal most MNs innervate fast muscle andare involved in large, fast swimming. During swimming,MNs are recruited from slow to intermediate to fast,and, therefore, from ventral MNs to dorsal MNs. Targetmuscle is not the only defining factor between thesegroups of neurons, as firing pattern, input resistance,reliability, and oscillatory drive, are just a few of the in-trinsic properties suspected to contribute to their differ-ential recruitment [14, 59, 60].Primary MNs, which innervate fast muscle, are known

to be responsible for the initial spontaneous coiling con-tractions and later escape behavior in zebrafish, whilevarious subsets of secondary MNs are necessary for allswim speeds. In a ned1 mutant where secondary MNsdegenerate, but primary MNs are preserved, normalspontaneous coiling contractions are present, but thefish cannot swim [33]. Although the purpose of theseseparate waves of neuronal birth remains elusive, somehypothesize primary MNs are necessary to form a basefor the development of locomotor CPG in the early em-bryonic spinal cord [19].Excitatory inputs onto axial MNs in zebrafish are pro-

vided by V2a interneurons defined by expression of theChx10 transcription factor [61–63]. It has been shownthat distinct V2a populations drive dorsal and ventraltrunk musculature in zebrafish [60, 64, 65]. Studies inboth zebrafish and lamprey disprove the previous notionthat only left-right alternation CPGs existed in primitiveaxial muscle control [64, 66]. This differential input con-tributes to the non-synchronous activation of thesemuscle groups important for behaviors such as posturalcontrol. Independent control of dorsal and ventral ipsi-lateral muscles is suggested to have been a template forseparate control of muscles on the same side of thebody, such as those in limbs [67].

Zebrafish are able to modulate their swimming speedthrough the recruitment of distinct MN subtypes. Whilethe MNs that drive different swimming speeds vary inanatomical size and excitability, studies suggest differen-tial recruitment of neurons along the dorso-ventral axisis not dependent solely on intrinsic properties but alsoon preferential excitatory drive [67]. Analogous to zebra-fish spinal MNs, interneurons are organized on thedorsal-ventral axis based on recruitment during swim-ming and birth order [62]. Dorsally positioned, earlyborn V2a neurons are active during higher frequencyswimming when ventral, late born V2a neurons areinhibited. At least for V2a neurons, the relation betweenposition and recruitment order does not persist into adultstages [14, 61, 68, 69]. However, experiments in adult zeb-rafish have revealed preferential connections and reliablemonosynaptic input from V2a neurons to proximal MNsrecruited at the same frequency of swimming, consistentwith the idea that different V2a neurons govern differentspeeds of locomotion [15, 61, 65, 69].While premotor inputs have a significant influence on

locomotor behavior, MNs are the ultimate gate to undu-lation in the zebrafish. Increasing evidence suggest MNsserve in an instructive manner to control the output oflocomotor circuits. A recent study demonstrated that, inaddition to having chemical synapses, some V2a inter-neurons in zebrafish are also electrically coupled to MNsvia gap junctions. This coupling permits the backwardpropagation of electrical signals from MNs influencingthe synaptic transmission and firing threshold of V2a in-terneurons, and therefore their recruitment during loco-motion [70]. These gap junctions allow the MNs tocontrol locomotor circuit function in a retrograde man-ner, causing the V2a interneurons and the MNs to act asa unit, which may contribute to the maintenance oflocomotor rhythm generation.

Functional diversity of axial motor circuits intetrapodsWhile a primary function of axial MNs is to drive loco-motion in zebrafish, in tetrapods MMC and HMC neu-rons play essential roles in multiple non-locomotorfunctions including breathing and maintaining spinalalignment. Some features of the locomotor CPG in fishappear to have been preserved in tetrapods to assist inlimb-based locomotion. For example, in amphibian andreptile species undulation of spinal segments can beused to facilitate movements of limbs [71]. In mammals,particularly in bipedal species, axial MNs appear to havebeen largely dissociated from locomotor CPG networks,which likely played an important role in enabling newtypes of axial muscle-driven motor behaviors.An important step in the evolution of axial motor cir-

cuits in tetrapods was the utilization of hypaxial muscle

D’Elia and Dasen Neural Development (2018) 13:10 Page 7 of 12

and its derivatives to support breathing on land. Expan-sion and contraction of the lungs during respiration ismediated by PMC and HMC neurons, which control thediaphragm and body wall muscle, respectively. In mam-mals, PMC and HMC firing is governed by CPG circuitslocated in the brainstem. Neurons in the preBötzinger(preBötz) complex and parafacial group provide the pre-dominant rhythmic drive to PMC and HMC neuronsduring inspiratory and expiratory breathing [57]. Brain-stem CPG networks target neurons in the ventral re-spiratory group (VRG) that in turn project to hypaxialand phrenic MNs within the spinal cord (Fig. 4a). Whilethe developmental logic that determines connectivity be-tween preBötz, VRG, and spinal MNs is not fully under-stood, a recent study has shown that connectivitybetween preBötz and VRG neurons rely on a commontranscription factor, Dbx1 [72]. Expression of Dbx1 isabsent from MNs, suggesting other intrinsic factors areinvolved in establishing connectivity between VRG andaxial MNs. Connections between brainstem respiratorycenters and spinal MNs could rely on actions ofsegmentally-restricted fate determinants, such Hoxgenes, which differentiate PMC and HMC from otherspinal MN subtypes (Fig. 4a) [73].While motor circuits controlling breathing and loco-

motion rely on rhythmically active neural circuits, the

development of motor circuits controlling posturalstabilization and spinal alignment have been more diffi-cult to study in mammals. In upright-walking bipedalvertebrates, the spine is kept in a relatively rigid config-uration. Studies in humans indicate that coactivation ofextensor and flexor axial muscles are essential for theload-bearing capacity and stability of the spine [74, 75].The circuits that stabilize spinal alignment are not wellcharacterized, but presumably require axial neural controlsystems that are fundamentally distinct from those con-trolling respiration in tetrapods and locomotion in fish.A recent study in mice has provided evidence that sen-

sory neurons play important roles in maintaining align-ment of the spine. Mutation in the Runx3 transcriptionfactor, which is required for the development of muscleproprioceptive sensory neurons (pSNs) [76], leads to aprogressive scoliosis of the spine (Fig. 4b-d) [77]. Thisphenotype does not appear to be a consequence of a re-quirement for Runx3 function in other tissues, sincesimilar results were observed after Runx3 deletion spe-cifically from pSNs. Although how this mutation affectsthe circuits involved in spinal stabilization is unclear, itis likely due to altered connections between pSNs andthe axial motor circuits essential for maintaining pos-ture. Loss and gain of function studies have shown thatRunx3 is required for the ability of pSNs to establish

a

b dc

Fig. 4 Diverse function of axial motor circuits in tetrapods. a Simplified diagram of respiratory networks for inspirational breathing. Rhythmgeneration in the preBötzinger (preBötz) complex is relayed to rostral ventral respiratory group (rVRG) neurons. rVRG neurons target PMC neuronsand HMC neurons in the spinal cord. Connections between preBötz and rVRG neuron relies on Dbx1 gene function. b-d Role of axial motorcircuits in spinal alignment. b Axial muscles and nerves associated with vertebrae. Box indicates region magnified in panel c. c Consequences ofRunx3 mutation on the projection of proprioceptive sensory neurons in the spinal cord. Loss of Runx3 leads to a loss of projections to MNs, andlikely other classes of spinal interneurons. d Effect of Runx3 mutation on vertebral alignment in adult mice

D’Elia and Dasen Neural Development (2018) 13:10 Page 8 of 12

connections with MNs and other neural classes [77–79],suggesting that the Runx3 mutant phenotype is due tothe disruption of local sensory-motor spinal reflex cir-cuits. In addition, mutations that affect the function ofthe MMC-restricted transcription factor Mecom alsocauses abnormal bending of the spine [80], raising thepossibility that this phenotype is also consequence of al-tered connectivity between axial MNs and premotorneural populations.

Developmental mechanisms of axial motor circuitassembly in tetrapodsThe distinct use of MMC neurons in locomotion andposture, while HMC and HMC-like MNs are essentialfor breathing, raise the question of how premotor cir-cuits dedicated to specific motor functions target the ap-propriate axial MN subtype. While the answer to thisquestion is largely unknown, studies characterizing thedistribution of spinal interneurons connected to specificMN columnar subtypes have provided a partial answer.Rabies-based monosynaptic tracing of interneurons con-nected to MMC and HMC neurons revealed that axialMNs receive local spinal premotor inputs that are evenlydistributed across both sides of the spinal cord (Fig. 5a).In contrast, limb MNs receive inputs predominantlyfrom premotor interneurons on the ipsilateral side of thespinal cord [81]. Axial MN dendritic arborization pat-terns are also distinct from those of limb MNs, whichmay help determine their specific connectivity with pre-motor interneuron populations (Fig. 5a). MMC neuronshave dendrites that extend across the midline, which ap-pears to enable them to capture a greater proportion ofinputs from contralateral interneuron populations, andestablish connectivity with interneurons distinct fromthose of HMC neurons. In contrast, limb-innervatingLMC neurons are found in more lateral and dorsal re-gions of the spinal cord and have radially-projectingdendrites, which may afford them greater input from ip-silateral interneuron populations.Do the molecular identities and/or positional differ-

ences between MN subtypes determine their premotorinput pattern and function? The ability to geneticallyalter the composition of MN subtypes within the mousespinal cord provides evidence that MN subtype identityplays an important role in determining the functionalproperties of spinal circuits. Conversion of limb MNs toan axial HMC fate, through deletion of limb MN deter-minant Foxp1, leads to the loss of limb-specific motoroutput patterns [82, 83]. In the absence of Foxp1, thenormal alternation of limb-flexor and -extensor firingpatterns is lost, and the remaining HMC-like popula-tions fire in a predominantly flexor-like pattern.Recent studies also indicate that determinants of MN

columnar identity play crucial roles in defining the

patterns and types of synaptic inputs that MNs receive[84]. Transformation of thoracic HMC neurons to alimb-level LMC fate, through mutation of the Hoxc9gene [85], shifts spinal premotor inputs to predomin-antly ipsilateral populations (Fig. 5b). In Hoxc9 mutants,the transformed HMC populations also settle in a moredorsolateral position, and their dendrites project radially,similar to those of limb-innervating MNs (Fig. 5b) [84].While these studies do not resolve the basic question ofhow differences between HMC and MMC inputs areachieved, they suggest that intrinsic differences betweenMN molecular identity, dendritic morphology, and pos-ition contribute to shaping the pattern of connection

a

b

Fig. 5 Developmental mechanisms of axial motor circuit assembly. aDendritic morphology and premotor input pattern for MN columnarsubtypes. MMC neurons have dendrites that extend across the midlineand their monosynaptic premotor inputs are distributed across bothsides of the spinal cord. Like MMC neurons, HMC neuron dendritesextend medio-laterally and have a similar premotor input distributionpattern. LMC neurons have radially organized dendrites and receivepremotor inputs predominantly from ipsilateral spinal interneurons.Darker shading indicates higher density of interneurons connected toMNs. b Effect of Hoxc9 mutation on premotor input pattern. In Hoxc9mutants, thoracic HMC neurons are converted to LMC fate, while MMCneurons are grossly unaffected. In Hoxc9 mutants, ectopic LMC neuronsstill project to intercostal muscle. The dendritic pattern of thoracic MNsin Hoxc9 mutants becomes more limb-like, and MNs projecting tointercostal muscle receive a higher distribution of inputs from ipsilateralpremotor interneurons. Diagram based on data in [84]

D’Elia and Dasen Neural Development (2018) 13:10 Page 9 of 12

within the motor circuits. How these genetic manipula-tions affect the function of axial motor circuits remainsto be determined. Nevertheless, analyses of Foxp1 andHoxc9 mutants indicate that the columnar identity ofspinal MNs plays a significant role in determining thearchitecture and output patterns of spinal circuits.

ConclusionsStudies on the development of neural circuits controllingaxial muscles have provided valuable insights into howspecific motor functions develop and have evolved in thevertebrate lineage. While we have a fairly in depth un-derstanding of the genetic programs controlling the spe-cification of tetrapod axial MN subtypes, how thesefunctionally diverse populations are connected to appro-priate higher order circuits remains to be determined.Recent studies showing that MN-intrinsic programscontribute to differences in the patterns of premotorconnectivity between limb and axial MNs suggests ageneral mechanism through which motor circuits are as-sembled, as a function of molecular differences in theirtarget MN populations. Further functional studies onthe consequences of disrupting MN differentiation couldprovide a means to test the role of MN subtype identityin the development of axial circuits essential for breath-ing and spinal alignment.Comparisons between species that use axial MNs for dis-

tinct functions have provided insights into how differentmotor behaviors are specified during development. Al-though this review has focused on vertebrate development,many of the intrinsic molecular features of axial MNs ap-pear to be conserved in invertebrates. Similar to verte-brates, in Drosophila and C. elegans subtypes of MNs canbe defined by expression of the transcription factors, Hb9,Lhx3, and Isl1/2 [86]. Since it is thought the ancestor to allbilaterians had a fairly complex nervous system [87, 88],and likely used an axial-like locomotor circuit to move, itwould be informative to know the extent to which theneural circuits governing axial muscle-driven locomotionhave been preserved across animal species.If an axial locomotor circuit represents the ancestral

condition in the common ancestor to bilaterians, thenwhat mechanisms were employed to generate the distinctneural circuits present in mammals? One example of howmotor circuits have changed is the use of axial muscle forlocomotion in fish, versus their non-locomotor functionsin tetrapods. Whether these differences reflect whole-salechanges in spinal circuits, or changes in a limited numberof circuit components remains to be determined. Furthercross-species comparisons of the functional roles of spe-cific interneuron and motor neuron subtypes will likelyprovide important clues into how axial motor circuits areestablished during development and have evolved acrossthe animal kingdom.

AbbreviationsCaP: Caudal primary motor neuron; CPG: Central pattern generator; dpf: Dayspost fertilization; dRoP: Dorsal rostral primary motor neuron; ei: Externalintercostal muscle; eo: External oblique muscle; ESC: Embryonic stem cell;FGF: Fibroblast growth factor; FGFR1: Fibroblast growth factor receptor 1;HMC: Hypaxial motor column; hpf: Hours post fertilization; ii: Internalintercostal muscle; ilio: Iliocostalis muscle; lc: Levator costae muscle;LMC: Lateral motor column; long: Longissimus muscle; MiP: Middle primarymotor neuron; MMC: Medial motor column; MN: Motor neuron;PGC: Preganglionic motor column; PMC: Phrenic motor column; pMN: Motorneuron progenitor; pSN: Proprioceptive sensory neuron; sc: Subcostalismuscle; Shh: Sonic hedgehog; sr: Caudal serratus muscle;tv: Transversospinalis muscle; VRG: Ventral respiratory group; vRoP: Ventralrostral primary motor neuron

AcknowledgementsWe thank David McLean and anonymous reviewers for feedback on the text.

FundingWork in the lab is supported by grants R01NS062822 and R01NS097550 fromthe NIH.

Authors’ contributionsJD and KD wrote the manuscript and prepared the figures. Both authorsread and approved the final manuscript.

Ethics approval and consent to participateNot applicable.

Competing interestsThe authors declare that they have no competing interests.

Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.

Received: 14 March 2018 Accepted: 26 April 2018

References1. Kiehn O. Decoding the organization of spinal circuits that control

locomotion. Nat Rev Neurosci. 2016;17:224–38.2. Catela C, Shin MM, Dasen JS. Assembly and function of spinal circuits for

motor control. Annu Rev Cell Dev Biol. 2015;31:669–98.3. Fetcho JR. A review of the organization and evolution of motoneurons

innervating the axial musculature of vertebrates. Brain Res. 1987;434:243–80.4. Romer AS, Parsons TS. The vertebrate body, vol. viii. Philadelphia: Saunders;

1977. p. 624.5. Romanes GJ. The motor pools of the spinal cord. Prog Brain Res. 1964;11:

93–119.6. Romanes GJ. The motor cell columns of the lumbo-sacral spinal cord of the

cat. J Comp Neurol. 1951;94:313–63.7. Landmesser L. The distribution of motoneurones supplying chick hind limb

muscles. J Physiol. 1978;284:371–89.8. Hollyday M, Jacobson RD. Location of motor pools innervating chick wing.

J Comp Neurol. 1990;302:575–88.9. Demireva EY, Shapiro LS, Jessell TM, Zampieri N. Motor neuron position and

topographic order imposed by beta- and gamma-catenin activities. Cell.2011;147:641–52.

10. Price SR, De Marco Garcia NV, Ranscht B, Jessell TM. Regulation of motorneuron pool sorting by differential expression of type II cadherins. Cell.2002;109:205–16.

11. Tsuchida T, Ensini M, Morton SB, Baldassare M, Edlund T, Jessell TM, Pfaff SL.Topographic organization of embryonic motor neurons defined by expressionof LIM homeobox genes. Cell. 1994;79:957–70.

12. Gutman CR, Ajmera MK, Hollyday M. Organization of motor pools supplyingaxial muscles in the chicken. Brain Res. 1993;609:129–36.

13. Smith CL, Hollyday M. The development and postnatal organization ofmotor nuclei in the rat thoracic spinal cord. J Comp Neurol. 1983;220:16–28.

D’Elia and Dasen Neural Development (2018) 13:10 Page 10 of 12

14. Ampatzis K, Song J, Ausborn J, El Manira A. Pattern of innervation andrecruitment of different classes of motoneurons in adult zebrafish. J Neurosci.2013;33:10875–86.

15. McLean DL, Fan J, Higashijima S, Hale ME, Fetcho JR. A topographic map ofrecruitment in spinal cord. Nature. 2007;446:71–5.

16. Liu DW, Westerfield M. Function of identified Motoneurones and coordinationof primary and secondary motor systems during Zebra fish swimming. J Physiol.1988;403:73–89.

17. Jung H, Baek M, D'Elia KP, Boisvert C, Currie PD, Tay BH, Venkatesh B, BrownSM, Heguy A, Schoppik D, Dasen JS. The ancient origins of neural substratesfor land walking. Cell. 2018;172:667–82. e615

18. Stifani N. Motor neurons and the generation of spinal motor neurondiversity. Front Cell Neurosci. 2014;8:293.

19. Dessaud E, McMahon AP, Briscoe J. Pattern formation in the vertebrateneural tube: a sonic hedgehog morphogen-regulated transcriptionalnetwork. Development. 2008;135:2489–503.

20. Jessell TM. Neuronal specification in the spinal cord: inductive signals andtranscriptional codes. Nat Rev Genet. 2000;1:20–9.

21. Dasen JS, De Camilli A, Wang B, Tucker PW, Jessell TM. Hox repertoires formotor neuron diversity and connectivity gated by a single accessory factor,FoxP1. Cell. 2008;134:304–16.

22. Sharma K, Sheng HZ, Lettieri K, Li H, Karavanov A, Potter S, Westphal H, PfaffSL. LIM homeodomain factors Lhx3 and Lhx4 assign subtype identities formotor neurons. Cell. 1998;95:817–28.

23. Sharma K, Leonard AE, Lettieri K, Pfaff SL. Genetic and epigenetic mechanismscontribute to motor neuron pathfinding. Nature. 2000;406:515–9.

24. Dasen JS, Jessell TM. Hox networks and the origins of motor neurondiversity. Curr Top Dev Biol. 2009;88:169–200.

25. Soundararajan P, Miles GB, Rubin LL, Brownstone RM, Rafuse VF. Motoneuronsderived from embryonic stem cells express transcription factors and developphenotypes characteristic of medial motor column neurons. J Neurosci. 2006;26:3256–68.

26. Peljto M, Dasen JS, Mazzoni EO, Jessell TM, Wichterle H. Functional diversityof ESC-derived motor neuron subtypes revealed through intraspinaltransplantation. Cell Stem Cell. 2010;7:355–66.

27. Agalliu D, Takada S, Agalliu I, McMahon AP, Jessell TM. Motor neurons withaxial muscle projections specified by Wnt4/5 signaling. Neuron. 2009;61:708–20.

28. Tan GC, Mazzoni EO, Wichterle H. Iterative role of notch signaling in spinalmotor neuron diversification. Cell Rep. 2016;16:907–16.

29. Tosney KW. Proximal tissues and patterned neurite outgrowth at thelumbosacral level of the Chick-embryo - partial and complete deletion ofthe somite. Dev Biol. 1988;127:266–86.

30. Tosney KW. Proximal tissues and patterned neurite outgrowth at thelumbosacral level of the Chick-embryo - deletion of the Dermamyotome.Dev Biol. 1987;122:540–58.

31. Shirasaki R, Lewcock JW, Lettieri K, Pfaff SL. FGF as a target-derivedchemoattractant for developing motor axons genetically programmed bythe LIM code. Neuron. 2006;50:841–53.

32. Saint-Amant L, Drapeau P. Time course of the development of motorbehaviors in the zebrafish embryo. J Neurobiol. 1998;37:622–32.

33. Drapeau P, Saint-Amant L, Buss RR, Chong M, McDearmid JR, Brustein E.Development of the locomotor network in zebrafish. Prog Neurobiol. 2002;68:85–111.

34. Moreno RL, Ribera AB. Zebrafish motor neuron subtypes differ electricallyprior to axonal outgrowth. J Neurophysiol. 2009;102:2477–84.

35. Myers PZ, Eisen JS, Westerfield M. Development and axonaloutgrowth of identified motoneurons in the zebrafish. J Neurosci.1986;6:2278–89.

36. Lewis KE, Eisen JS. From cells to circuits: development of the zebrafishspinal cord. Prog Neurobiol. 2003;69:419–49.

37. Appel B, Korzh V, Glasgow E, Thor S, Edlund T, Dawid IB, Eisen JS.Motoneuron fate specification revealed by patterned LIM homeoboxgene expression in embryonic zebrafish. Development. 1995;121:4117–25.

38. Seredick SD, Van Ryswyk L, Hutchinson SA, Eisen JS. Zebrafish Mnx proteinsspecify one motoneuron subtype and suppress acquisition of interneuroncharacteristics. Neural Dev. 2012;7:35.

39. Park HC, Mehta A, Richardson JS, Appel B. olig2 is required for zebrafishprimary motor neuron and oligodendrocyte development. Dev Biol. 2002;248:356–68.

40. Hutchinson SA, Cheesman SE, Hale LA, Boone JQ, Eisen JS. Nkx6 proteinsspecify one zebrafish primary motoneuron subtype by regulating late islet1expression. Development. 2007;134:1671–7.

41. Seredick S, Hutchinson SA, Van Ryswyk L, Talbot JC, Eisen JS. Lhx3 and Lhx4suppress Kolmer-Agduhr interneuron characteristics within zebrafish axialmotoneurons. Development. 2014;141:3900–9.

42. Hutchinson SA, Eisen JS. Islet1 and Islet2 have equivalent abilities to promotemotoneuron formation and to specify motoneuron subtype identity.Development. 2006;133:2137–47.

43. Jung H, Dasen JS. Evolution of patterning systems and circuit elements forlocomotion. Dev Cell. 2015;32:408–22.

44. Beattie CE, Hatta K, Halpern ME, Liu HB, Eisen JS, Kimmel CB. Temporalseparation in the specification of primary and secondary motoneurons inzebrafish. Dev Biol. 1997;187:171–82.

45. Menelaou E, McLean DL. A gradient in endogenous rhythmicity andoscillatory drive matches recruitment order in an axial motor pool.J Neurosci. 2012;32:10925–39.

46. Rousso DL, Gaber ZB, Wellik D, Morrisey EE, Novitch BG. Coordinated actionsof the forkhead protein Foxp1 and Hox proteins in the columnarorganization of spinal motor neurons. Neuron. 2008;59:226–40.

47. Catela C, Shin MM, Lee DH, Liu JP, Dasen JS. Hox proteins coordinate motorneuron differentiation and connectivity programs through ret/Gfr alphagenes. Cell Rep. 2016;14:1901–15.

48. Lacombe J, Hanley O, Jung H, Philippidou P, Surmeli G, Grinstein J, DasenJS. Genetic and functional modularity of Hox activities in the specification oflimb-innervating motor neurons. PLoS Genet. 2013;9:e1003184.

49. Mendelsohn AI, Dasen JS, Jessell TM. Divergent Hox coding and evasion ofretinoid signaling specifies motor neurons innervating digit muscles.Neuron. 2017;93:792–805. e794

50. Philippidou P, Walsh CM, Aubin J, Jeannotte L, Dasen JS. Sustained Hox5gene activity is required for respiratory motor neuron development. NatNeurosci. 2012;15:1636–44.

51. Landry-Truchon K, Fournier S, Houde N, Rousseau JP, Jeannotte L, KinkeadR. Respiratory consequences of targeted losses of Hoxa5 gene function inmice. J Exp Biol. 2017;220:4571–7.

52. Bermingham JR Jr, Scherer SS, O'Connell S, Arroyo E, Kalla KA, Powell FL,Rosenfeld MG. Tst-1/Oct-6/SCIP regulates a unique step in peripheralmyelination and is required for normal respiration. Genes Dev. 1996;10:1751–62.

53. Machado CB, Kanning KC, Kreis P, Stevenson D, Crossley M, Nowak M,Iacovino M, Kyba M, Chambers D, Blanc E, Lieberam I. Reconstruction ofphrenic neuron identity in embryonic stem cell-derived motor neurons.Development. 2014;141:784–94.

54. Hanley O, Zewdu R, Cohen LJ, Jung H, Lacombe J, Philippidou P, Lee DH,Selleri L, Dasen JS. Parallel Pbx-dependent pathways govern thecoalescence and fate of motor columns. Neuron. 2016;91:1005–20.

55. Moens CB, Selleri L. Hox cofactors in vertebrate development. Dev Biol.2006;291:193–206.

56. Grillner S, Jessell TM. Measured motion: searching for simplicity in spinallocomotor networks. Curr Opin Neurobiol. 2009;19:572–86.

57. Feldman JL, Del Negro CA, Gray PA. Understanding the rhythm ofbreathing: so near, yet so far. Ann Rev Physiol. 2013;75(75):423–52.

58. Grillner S. Biological pattern generation: the cellular and computationallogic of networks in motion. Neuron. 2006;52:751–66.

59. Wang WC, Brehm P. A gradient in synaptic strength and plasticity amongMotoneurons provides a peripheral mechanism for locomotor control. CurrBiol. 2017;27:415–22.

60. Menelaou E, VanDunk C, McLean DL. Differences in the morphology ofspinal V2a neurons reflect their recruitment order during swimming in larvalzebrafish. J Comp Neurol. 2014;522:1232–48.

61. Eklof-Ljunggren E, Haupt S, Ausborn J, Dehnisch I, Uhlen P, Higashijima S, ElManira A. Origin of excitation underlying locomotion in the spinal circuit ofzebrafish. Proc Natl Acad Sci U S A. 2012;109:5511–6.

62. Kimura Y, Okamura Y, Higashijima S. Alx, a zebrafish homolog of Chx10,marks ipsilateral descending excitatory interneurons that participate in theregulation of spinal locomotor circuits. J Neurosci. 2006;26:5684–97.

63. Ljunggren EE, Haupt S, Ausborn J, Ampatzis K, El Manira A. Optogeneticactivation of excitatory premotor interneurons is sufficient to generatecoordinated locomotor activity in larval zebrafish. J Neurosci. 2014;34:134–9.

64. Bagnall MW, McLean DL. Modular organization of axial microcircuits inzebrafish. Science. 2014;343:197–200.

D’Elia and Dasen Neural Development (2018) 13:10 Page 11 of 12

65. McLean DL, Masino MA, Koh IY, Lindquist WB, Fetcho JR. Continuous shiftsin the active set of spinal interneurons during changes in locomotor speed.Nat Neurosci. 2008;11:1419–29.

66. Wallen P, Grillner S, Feldman JL, Bergelt S. Dorsal and ventral myotomeMotoneurons and their input during fictive locomotion in lamprey. J Neurosci.1985;5:654–61.

67. McLean DL, Dougherty KJ. Peeling back the layers of locomotor control inthe spinal cord. Curr Opin Neurobiol. 2015;33:63–70.

68. Gabriel JP, Ausborn J, Ampatzis K, Mahmood R, Eklof-Ljunggren E, El ManiraA. Principles governing recruitment of motoneurons during swimming inzebrafish. Nat Neurosci. 2011;14:93–9.

69. Ampatzis K, Song J, Ausborn J, El Manira A. Separate microcircuit modulesof distinct v2a interneurons and motoneurons control the speed oflocomotion. Neuron. 2014;83:934–43.

70. Song J, Ampatzis K, Bjornfors ER, El Manira A. Motor neurons controllocomotor circuit function retrogradely via gap junctions. Nature. 2016;529:399–402.

71. Chevallier S, Ijspeert AJ, Ryczko D, Nagy F, Cabelguen JM. Organisation ofthe spinal central. Pattern generators for locomotion in the salamander:biology and modelling. Brain Res Rev. 2008;57:147–61.

72. Wu J, Capelli P, Bouvier J, Goulding M, Arber S, Fortin G. A V0 core neuronalcircuit for inspiration. Nat Commun. 2017;8:544.

73. Philippidou P, Dasen JS. Hox genes: choreographers in neural development,architects of circuit organization. Neuron. 2013;80:12–34.

74. Crisco JJ, Panjabi MM, Yamamoto I, Oxland TR. Euler stability of the humanligamentous lumbar spine. Part II: experiment. Clin Biomech (Bristol, Avon).1992;7:27–32.

75. Cholewicki J, Panjabi MM, Khachatryan A. Stabilizing function of trunkflexor-extensor muscles around a neutral spine posture. Spine (Phila Pa1976). 1997;22:2207–12.

76. Levanon D, Bettoun D, Harris-Cerruti C, Woolf E, Negreanu V, Eilam R,Bernstein Y, Goldenberg D, Xiao C, Fliegauf M, Kremer E, Otto F, Brenner O,Lev-Tov A, Groner Y. The Runx3 transcription factor regulates developmentand survival of TrkC dorsal root ganglia neurons. EMBO J. 2002;21:3454–63.

77. Blecher R, Krief S, Galili T, Biton IE, Stern T, Assaraf E, Levanon D, Appel E,Anekstein Y, Agar G, Groner Y, Zelzer E. The proprioceptive systemmasterminds spinal alignment: insight into the mechanism of scoliosis. DevCell. 2017;42:388–99. e383

78. Chen AI, de Nooij JC, Jessell TM. Graded activity of transcription factorRunx3 specifies the laminar termination pattern of sensory axons in thedeveloping spinal cord. Neuron. 2006;49:395–408.

79. Kramer I, Sigrist M, de Nooij JC, Taniuchi I, Jessell TM, Arber S. A role forRunx transcription factor signaling in dorsal root ganglion sensory neurondiversification. Neuron. 2006;49:379–93.

80. Juneja SC, Vonica A, Zeiss C, Lezon-Geyda K, Yatsula B, Sell DR, Monnier VM,Lin S, Ardito T, Eyre D, Reynolds D, Yao Z, Awad HA, Yu H, Wilson M,Honnons S, Boyce BF, Xing L, Zhang Y, Perkins AS. Deletion of Mecom inmouse results in early-onset spinal deformity and osteopenia. Bone. 2014;60:148–61.

81. Goetz C, Pivetta C, Arber S. Distinct limb and trunk premotor circuitsestablish laterality in the spinal cord. Neuron. 2015;85:131–44.

82. Machado TA, Pnevmatikakis E, Paninski L, Jessell TM, Miri A. Primacy offlexor locomotor pattern revealed by ancestral reversion of motor neuronidentity. Cell. 2015;162:338–50.

83. Hinckley CA, Alaynick WA, Gallarda BW, Hayashi M, Hilde KL, Driscoll SP,Dekker JD, Tucker HO, Sharpee TO, Pfaff SL. Spinal locomotor circuitsdevelop using hierarchical rules based on Motorneuron position andidentity. Neuron. 2015;87:1008–21.

84. Baek M, Pivetta C, Liu JP, Arber S, Dasen JS. Columnar-intrinsic cues shapepremotor input specificity in locomotor circuits. Cell Rep. 2017;21:867–77.

85. Jung H, Lacombe J, Mazzoni EO, Liem KF Jr, Grinstein J, Mahony S,Mukhopadhyay D, Gifford DK, Young RA, Anderson KV, Wichterle H, DasenJS. Global control of motor neuron topography mediated by the repressiveactions of a single hox gene. Neuron. 2010;67:781–96.

86. Thor S, Thomas JB. Motor neuron specification in worms, flies and mice:conserved and 'lost' mechanisms. Curr Opin Genet Dev. 2002;12:558–64.

87. De Robertis EM, Sasai Y. A common plan for dorsoventral patterning inBilateria. Nature. 1996;380:37–40.

88. Arendt D, Denes AS, Jekely G, Tessmar-Raible K. The evolution of nervoussystem centralization. Philos Trans R Soc Lond Ser B Biol Sci. 2008;363:1523–8.

D’Elia and Dasen Neural Development (2018) 13:10 Page 12 of 12